Capturing Luminous Flux Entering Human Eyes with a Camera, Part 2: A Field Verification Experiment

ABSTRACT

A space occupant’s real-time exposure to light reaching his/her eyes has been a concern in lighting research and design practice. Following the parallel fundamental research in Part 1, Part 2 of this study explored a way to use a Canon 550D fitted with the Sigma 4.5 mm F2.8 circular fisheye lens to estimate the luminous flux arriving at the entrance pupil from the visible luminous environment. By taking two 180° HDR images at the same camera position but in different aiming directions and merging them into a single luminance map covering the observer’s entire binocular visual field (214° horizontally and 125° vertically), the monocular and binocular luminance map can be generated with the aid of the dynamic visual field model proposed in Part 1. In a field verification experiment, the luminous flux passing through the entrance pupil in the preset scenes and gaze angles was estimated. Individual luminous contribution from an object or a region of interest within the occupant’s dynamic binocular visual field or subfields was also calculated. This study expanded the capacity of HDR photography in lighting measurement to capture binocular luminous flux entering human eyes as input for optimizing human-light interaction in various luminous environments.

KEYWORDS:

• Gaze point
• dynamic visual field
• high dynamic range imaging

Nomenclature

$\left(x_{i1},y_{i1}\right)$	=	Image pixel coordinate before the \transformation
$\left(x_{i2},y_{i2}\right)$	=	Image pixel coordinate after the transformation
$\left(X_{i1},Y_{i1},Z_{i1}\right)$	=	Spatial coordinate before the transformation
$\left(X_{i2},Y_{i2},Z_{i2}\right)$	=	Spatial coordinate after the transformation
$\kappa$	=	Yaw angle
$\eta$	=	Pitch angle
$R_z\left(\kappa \right)$	=	Rotation matrix along the yaw angle
$R_x\left(\eta \right)$	=	Rotation matrix along the pitch angle
$\mathrm{\angle }R$	=	The calibrated real incident angle
$\mathrm{\angle }C$	=	The incident angle retrieving from the circular fisheye lens image.
${\kappa }_{cam}$	=	Yaw angle of the camera rotation
$\kappa 1_{cam},\kappa 2_{cam}$	=	The first and second yaw angles of the camera rotation
${\eta }_{cam}$	=	Pitch angle of the camera rotation
$A_{t\mathrm{\_}img},A_i$	=	The projected area of the target and pixel in the image
$A{\prime }_t$	=	The apparent area of a target
${\mathrm{\Omega }}_t$	=	The solid angle of a single target
$N_t$	=	The pixel number of the target
${\mathrm{\Omega }}_i$	=	The per-pixel solid angle
$d$	=	The distance from a target to the camera
${\theta }_{Gaze}$	=	Incident angle of visual axis
${\phi }_{Gaze}$	=	Azimuth angle of visual axis

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

Funding Siqi He was a visiting Ph.D. student at the University of Kansas Lighting Research Laboratory, where the research work was conducted. Siqi was supported by the Chinese Scholarship Council of the Ministry of Education (No.201906050075).